Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/mikuare/code-ledger/codeledgergit clone --depth 1 https://github.com/mikuare/Code-LedgerWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00018 | $0.00661 |
| Opus 5 | $0.00009 | $0.00331 |
| Sonnet 5 | $0.00004 | $0.00132 |
| Haiku 4.5 | $0.00002 | $0.00066 |
Grade A, and why
codeledger scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
CodeLedger is this project's local engineering memory. Route on $ARGUMENTS:
- empty, or
check— runcodeledger check. Report the verdict line, then any component that is notOK, then the recommended commands verbatim.STALEmeans the index is behind the working tree and every other answer describes an earlier state; say so plainly rather than reporting it as a minor detail. check --deep— runcodeledger check --deep. Same, and it also launches the MCP server and completes a real handshake.map, ormap <area>— runcodeledger map(adding the area if one was given). Summarise the areas, the entry points and the most-depended-on symbols. This is what the project looks like from the index; you did not need to open a file for it.doctor— runcodeledger doctorand report what it found.guard— runcodeledger guard. Report whether anything that was working is gone, with each finding's options; never pick an option yourself.value— runcodeledger valueand summarise what CodeLedger has recorded doing here.done <task>— runcodeledger done "<task>" --agent claude-code. Report the verdict and, below DONE, exactly what remains; do not soften it.- anything else — treat
$ARGUMENTSas the task and runcodeledger brief "$ARGUMENTS". Do not rewrite the task first; pass it exactly as given.
For brief, act on do_next:
ASK_USER— stop and ask the user the questions it returned, with their options. Do not pick one yourself, and do not start work on the most plausible reading. The whole point of the question is that the repository cannot answer it.CONFIRM— restate the task in your own words (what changes, what does not, how it is verified) and runcodeledger brief "$ARGUMENTS" --understanding "<your restatement>". ReportCONFIRMEDorMISMATCHbefore editing.RESUME— reportnext_actionandfailed_attemptsfirst. Do not repeat an approach already recorded as failed.RECONSIDER— say what the previous attempts actually changed, and why repeating them is unlikely to help.PROCEED_WITH_CARE/PROCEED— report the target, the files in scope and the measured footprint, then continue with the task.
Report what CodeLedger said. Where it says UNKNOWN, say UNKNOWN — it means the evidence is absent, never that there is nothing to worry about.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 27 lines · 18 tokens per session scan A 21b6cb701156
codeledger is a command published in the GitHub repository mikuare/Code-Ledger (0 stars, last pushed 6d ago), licensed MIT. It adds 18 tokens to every session and 661 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
checklist
Generate a custom checklist for the current feature based on user requirements.
clarify
Identify underspecified areas in the current feature spec by asking up to 5 highly targeted clarification questions and encoding answers back into the spec.
specify
Create or update the feature specification from a natural language feature description.
analyze
Perform a non-destructive cross-artifact consistency and quality analysis across spec.md, plan.md, and tasks.md after task generation.
converge
Assess the current codebase against the feature's spec, plan, and tasks, then append any remaining unbuilt work as new tasks to tasks.md so implement can complete it.
implement
Execute the implementation plan by processing and executing all tasks defined in tasks.md.